Abstract
Abstract When planning a trip, people are searching for information. They want to know what choices they might have and they must evaluate these choices in order to choose the best of them. If they are going to a traditional travel agency, the agent will give them advises, based on personal knowledge or other customers’ experience. Therefore when we develop a trip planning system, we must know how to model user’s preference. As users differ in many parameters, like status, expertise and preferences, all these should be taken into account to improve the quality of trip planning. Visiting spots also have many constraints such as opening time, closing time and cost. The trip planning system has to satisfy not only the customer’s personal preferences and constraint, but also the complex spatial, temporal, physical constraints and cost imposed by the transportation methods among visiting spots and various environmental supports such as the availability of hotels. In this thesis, we are in particular interested in understanding how a travel agent could conduct such kinds of recommendation and negotiation dialogues with the user and come out with a satisfactory trip plan for the user. We use case-based reasoning to plan personal route. Besides, we will recommend visiting spots for users according to users’ experiences. We will focus on how to resolve of constraint violation when there are conflicts between user’s preferences and visiting spots’ time constraints.